High hyperdiploidy accounts for 20–30% of childhood ALL. There is a distinct pattern of chromosomes gained at each modal number of chromosomes: most commonly gained is chromosome 21 (ch21), followed by ch4, chX, ch10, ch6, ch14, ch18, and ch17 [27]. Previous clinical studies have demonstrated that certain combination of gained chromosome are associated with better prognosis: “double trisomy” of ch4 and ch10 in the Pediatric Oncology Group (POG); “triple trisomy” of ch4, ch10, and ch17 in Children’s Oncology Group (COG); trisomy of ch11 and ch17 in the Tokyo Children’s Cancer Study Group (TCCSG) [27–29]. Although prognostic significance of specific gained chromosomal combination is inconsistent among different study groups, high hyperdiploidy itself is associated with CD10-positive B-cell precursor phenotype, low leukocyte (WBC) count, younger age at diagnosis, and excellent prognosis.

In contrast, hypodiploidy accounts for only 1% of childhood ALL and is a strong predictor of poor prognosis [30, 31]. Recent genomic analysis has identified TP53 mutations in 91% of low-hypodiploid ALL (32–39 chromosomes) [32]. Surprisingly, the mutation was also present in germline in 43% of the cases, which is a hallmark of Li-Fraumeni syndrome. Occasionally, there are cases with “masked hypodiploidy,” which occurs as a result of doubling of a hypodiploid clone [33]. Caution is needed because they do not appear to be different from non-masked hypodiploidy in clinical and prognostic features but could be misdiagnosed as “hyperdiploidy” harboring opposite prognosis.

t(12;21)(p13;q22.1) accounts for 15–20% of childhood ALL. It is cryptic in most cases and could be detected by FISH or real-time quantitative polymerase chain reaction (PCR). The translocation results in fusion of two hematopoietic transcription factor genes, ETV6 (formerly known as TEL) and RUNX1 (AML1). ETV6-RUNX1 appears to arise in utero, but the fusion solely itself is not sufficient to cause the leukemia and subsequent events are required to full progression [37]. Children with ETV6-RUNX1 ALL is associated with excellent outcome similar to high hyperdiploid ALL [38].

t(1;19)(q23;p13.3) is the second most common translocation and accounts for 5% of childhood ALL. The translocation results in the fusion of two transcription factor genes, TCF3 (formerly known as E2A) and PBX1. TCF3-PBX1 ALL appears to be pre-B cell phenotype with positive cytoplasmic μ. It was once associated with poor prognosis but has lost its prognostic significance in the context of modern ALL chemotherapy especially that contains high-dose methotrexate [39]. However, TCF3-PBX1 is associated with higher risk of central nervous system (CNS) relapse [40].

There is a rare subtype of TCF3 gene involved ALL with fusion partner hepatic leukemia factor (HLF) gene which arise from t(17;19)(q22;p13). TCF3-HLF only occurs in less than 1% of childhood ALL but has unique characteristics such as hypercalcemia, an increased risk of disseminated intravascular coagulation (DIC), and, moreover, very poor prognosis [41].

KMT2A (MLL) gene, located at chromosome band 11q23, encodes a histone methyltransferase that is involved in epigenetic regulation of blood cell development via expression of multiple Hox genes [42]. Rearrangements of MLL occur as a result of balanced chromosomal translocations that fuse the MLL gene to one of more than 70 known partner genes. The most common in ALL is MLL-AF4 (KMT2A-AFF1) derived from t(4;11)(q21;q23), followed by MLL-ENL (KMT2A-MLLT1) from t(11:19)(q23;p13), and MLL-AF9 (KMT2A-MLLT3) from t(9;11)(p22;q23) [43]. MLL rearrangement is particularly common in ALL in infants (age <1 year old), which accounts for nearly 80% of the cases and associated with CD10-negative pro-B cell phenotype, high leukocyte count at diagnosis, and very poor prognosis (Table 2.2) [44–48]. ALL with MLL rearrangements have very few additional somatic mutations, one of the lowest of any sequenced cancers [49].

t(9;22)(q34;q11.2) results in formation of Philadelphia (Ph) chromosome, which is one of the first leukemic translocations described. Ph encodes BCR-ABL1, an activated tyrosine kinase. It is the most common translocation in adult ALL (25% of the cases) but less common in children (less than 5% of the cases). Ph-positive (Ph+) ALL was one of the most difficult to cure ALL with only 30–40% event-free survival (EFS) rate despite intensive chemotherapy and allogeneic hematopoietic stem cell transplantation [50]. However, introduction of imatinib, a selective tyrosine kinase inhibitor (TKI), has revolutionized the therapy for Ph+ ALL (Table 2.3) [51–54].

Ph-like or BCR-ABL1-like ALL is a newly recognized entity, harboring similar gene expression profile to Ph+ ALL but without BCR-ABL1 fusion gene [55, 56]. Recent genomic analyses have revealed diverse range of genetic alterations that activate tyrosine kinase signaling in 90% of the cases. The most commonly identified alterations are rearrangements involving ABL1, ABL2, CRLF2, CSF1R, EPOR, JAK2, NTRK3, PDGFRB, PTK2B, TSLP, or TYK2 and sequence mutations involving FLT3, IL7R, or SH2B3 [57]. Importantly, “ABL-class” kinases (ABL1, ABL2, CSF1R, and PDGFRβ) could be targeted with TKI such as imatinib and dasatinib and alterations that activate JAK-STAT signaling (JAK1, JAK2, JAK3, CRLF2, EPOR, TSLP, and IL7R) with JAK inhibitors [58].

ALL with intrachromosomal amplification of chromosome 21 (iAMP21) is characterized by amplification of a portion of ch21, which could be detected by FISH analysis using a probe for the RUNX1 gene that reveals five or more copies of the gene (or three extra copies on a single abnormal ch21 in metaphase FISH) [59]. It occurs in 2% of childhood ALL and is associated with poor prognosis [60].

Most commonly targeted genes involved in B-lymphoid development are PAX5 and IKZF1 that are mutated in 31% and 15% of children with B-ALL, respectively, [61]. Notably, recurrent deletions and inactivating mutations in IKZF1, which encodes the hematopoietic transcription factor IKAROS, are associated with very poor prognosis in B-ALL [56]. Genetic alterations of IKZF1 occur more frequently in high-risk cases, including Ph+ ALL and Ph-like ALL [62].

Cytokine receptor-like factor 2 (CRLF2) forms a heterodimeric cytokine receptor with interleukin-7 receptor α (IL7Rα) that mediates B-cell precursor proliferation and survival via activation of downstream JAK/STAT pathways. Rearrangements of CRLF2 gene as IGHa-CRLF2 or P2RY8-CRF2 result in CRLF2 overexpression and found in 5–8% of pediatric ALL, 10–15% of adult ALL, and more than 50% of Down syndrome ALL [63–65]. It has been reported as poor prognostic factor but still controversial.

JAK2 mutations are found in 18–35% of Down syndrome ALL cases and 11% of non-Down high-risk BCR-ABL1-negative ALL cases [66]. The mutation cause constitutive JAK-STAT activation. There is overlap among IKZF1 and CRLF2 alterations [67].

In contrast to genetic alterations discovered in B-ALL, many of the genetic alterations in T-ALL have not been found to have prognostic value yet. However, one subset of T-ALL with unique biology, early T-cell precursor (ETP) ALL that accounts for 10–15% of pediatric T-ALL, is recognized as a new entity [68]. ETP-ALL was originally identified by its unique gene expression pattern and immunophenotype with very early T-cell progenitor features: absence of T-cell markers CD1a and CD8; dim or absent CD5; combined with cytoplasmic, but not surface, CD3; and positive for one or more myeloid/stem cell markers (CD34, CD117, HLA-DR, CD13, CD33, CD11b, or CD65). By whole-genome sequencing, gene mutations similar to myeloid leukemias including RAS signal pathways were identified [69].

Genomic studies of matched diagnosis and relapsed ALL sample pairs have revealed that only 42% of the cases had evolved from diagnosis clone (8% was same and 34% was clonal evolution from diagnosis clone), but majority (52%) was rather evolved from ancestral clones, and the remaining 6% was genetically distinct secondary leukemia [70]. By further sequencing studies, mutations including CREBBP, TP53, and NT5C2 were identified [71–74]. Focal deletion and sequence mutations in CREBBP were found in 19% of children with relapsed B-ALL. The mutation causes impaired histone acetylation and transcriptional regulation of CREBBP targets, thus impairing the normal CREBBP-mediated transcriptional response to glucocorticoids and possibly results in steroid therapy resistance of the relapsed clone. NT5C2 encodes as 5′-nucleotidase enzyme that is involved in metabolisms of 6-mercaputopurine. Its mutation causes increased enzyme activity and resistance to nucleoside analog therapy. TP53 alterations were identified in 12% of relapsed B-ALL and 6% of relapsed T-ALL cases in the Berlin-Frankfurt-Münster (BFM) study, which was enriched compared to the initial diagnosis samples and was associated with poor response to the therapy and inferior outcome.

Prognostic factors are the factors that are predictive of disease outcome and are usually used to tailor therapy (“risk-stratified therapy”) with the intention to minimize toxic events for patients with better prognosis and to maximize the treatment effect for patients with poorer prognosis. The currently well-recognized prognostic factors could be categorized into three groups.